3
\$\begingroup\$

Recently I asked an opinion based question regarding LED strips. I have learned my lesson and now I have an actual question regarding "how power flows".

I'll explain: I have a lighting project that has 6,6 meters of LED strip and power is injected at the beginning of the strip and at the end of the strip, it goes in an almost complete loop. That long piece of LED strip has a maximum draw of 10,xx A @ 12 volts, but for simplicity let's say 10A. I never intend to drive the LED's at full brightness, so unless I code a bug in my controllers software, it should never exceed 10A anyway. Now for the actual question:

If I fuse the wiring so that there is one 10A fuse right after the PSU and 5A fuses right before injecting the power to the strip at both injection spots, is that conceptually correct? My idea is that half of the strip would be powered on each end of the strip and I assume it would be safer.

Does the situation change if I cut the strip at the middle point and only connect data and ground at that point, and power both halves individually?q&d drawing of the circuit

  • Yellow=data, red=12v, black=ground
  • If necessary, I'll provide a better picture

Edit1: Or if someone can suggest a better solution altogether. I am being very neurotic about the safety of this installation.

Edit2: I thought about it a little while today and I have a hunch the below version would be better when it comes to being sure only 5A goes through each splice, maybe I could even get away with using smaller wire safely this way?rev2

As I mentioned, if anyone has a better way to achieve this result, I'm open to ideas.

\$\endgroup\$
4
  • \$\begingroup\$ What type of fuse? How quickly would it trip? Multiple fuses in series may prevent tripping some fuses (hopefully not the kind you will use) Do you know why people are not supposed to daisy chain fused power strips? Some psu will come with such a fuse on the wall AC line-. \$\endgroup\$
    – Abel
    Aug 14, 2022 at 11:15
  • \$\begingroup\$ I was planning on using the kind used in cars (plate fuse?), I'm not native English speaker and do not know what they are called. And to be completely honest, I do not know why that is an issue, luckily I asked here before doing anything :D edit: If you understand what I am trying to achieve here, what would be the correct way to do it? \$\endgroup\$ Aug 14, 2022 at 11:22
  • \$\begingroup\$ Fuse after PS not needed if you choose properly wire size. PS should have over current protection. Wire should hold the biggest current. In common you should decide what will be protected: wires, LEDs, PS. \$\endgroup\$
    – user263983
    Aug 14, 2022 at 16:07
  • \$\begingroup\$ Yeah, it has protection. So you reckon the 5A's would be enough? Conceptually the 10A fuse was meant to be a failsafe, sort of a bug detector rather than "actual fuse". \$\endgroup\$ Aug 14, 2022 at 18:46

2 Answers 2

2
\$\begingroup\$

Wiring fuses are used to prevent wires from overheating and melting insulation or starting a fire. In that case, the key thing is that the wire diameter be large enough to conduct the maximum current the fuse will pass. Copper wire will not melt until it gets to a very, very hot and unsafe temperature. So the focus here is on the electrical insulation around the wire. The maximum current is obtained from an "ampacity chart." The ampacity depends on the temperature rating of insulation on the wire.

For installations subject to the National Electric Code (NEC) you can use an NEC ampacity table. Outside the USA I am sure there is a similar table. But in your case, they don't cover small wires and small loads, so it won't be much use.

I found a table online here: https://www.southwire.com/medias/sys_master/pdfs/pdfs/hb7/hc9/8872277901342/SW-1003267-Lead-Wire-Maximum-Current-Capacity-Whitepaper-LO-1-.pdf

It is in a whitepaper published by Southwire Corporation. The title is "Lead Wire Maximum Current Capacity."

enter image description here

Looking at the table, I would say you are safe using 18 AWG wire with 5 or 10 Amp fuses as long as the wire is not routed through some type of thermal insulation or something that prevents it from cooling down.

You should read the whole paper. It has more than just the ampacity chart.

Another important concept is providing some margin. If your expected load is 5 Amps, you would usually not use a 5 A fuse. You would use a slightly larger fuse, perhaps 7.5 Amps, to prevent nuisance blowing of the fuse. In that case you would want to make sure the wire can carry 7.5 Amps without over-heating. The fuse and wire always must be considered together.

If you are worried about faults in the strip itself, it would be best to be explicit about what fault you are contemplating and what kind of failures are acceptable. For sure a simple short-circuit will blow a 5 or 7.5 A fuse (assuming the power supply does not have short-circuit current limiting or protection).

Device designers often put fuses inside products to mitigate specific failure scenarios. But the details can vary quite a bit, and would likely be tested by simulating or inducing the specific failure first without a fuse, and then adding the fuse if needed. For example, if you short out a power rail capacitor, what happens? If a trace overheats and starts to smoke, you might choose to put a fuse somewhere in series with the power input for the regulator to prevent that from happening. But that is all very device-specific.

So in summary, wiring fuses are based on current carried by the wire and the wire diameter. Device fuses are based on specific knowledge of the device and specific failure scenarios, and will usually involve testing of some sort. For best user experience, fuses should be de-rated to make sure they do not blow during normal operation. However, the wiring diameter must be based on the nominal fuse rating, not the normal operating current.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ The three most obvious failures for you would be power-to ground short, LED fails open-circuit, and one or more LEDs fail short circuit. You could actually test all of them if you like and see what happens. However, in order to do testing like that safely, you have to make sure you can easily and remotely de-energize the whole test setup, and that you can easily get away from the test setup if it starts to smoke (so you don't breathe smoke), and you have to make sure no fire can spread if the device under test should catch fire. That is just a sketch of safety requirements. \$\endgroup\$
    – user57037
    Aug 14, 2022 at 20:02
0
\$\begingroup\$

If the cable lengths are symmetric the setup will work. Cable resistance will rise with temperature, which will help in balancing the current.

Generally the fuses should be sized to prevent fire in failure modes, which is difficult to achieve in this case as a (semi-)shorted LED would present a hazard even if the current stays within rating. Undersized fuses (5 A vs 10 A) on both ends will protect from the situation where one power line is cut.

Cables should in this case be sized for 10 A, although it's quite certain that the LED strip itself is going to be where the overheating occurs.

Check the fuse datasheets for tripping time on rated current if you expect to get close to the rating.

\$\endgroup\$
1
  • \$\begingroup\$ Unfortunately, out of necessity, cable lengths are not equal, at strip start, there's only a few inches and other end is 3 feet away from PSU. What if I cut the strip so that there are 2 individual 3,3 meter strips, which are only connected to data and common ground, 12v individually? Also I guess I could just add 3 feet of cable at the start an coil it in the box \$\endgroup\$ Aug 14, 2022 at 11:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.